Opinion丨DTC transformation of the new economy, a simple and powerful data platform is crucial

It has been two years since the epidemic lasted, and a lot has changed. I believe everyone has a deep understanding of this point. The most direct point is that people go out less, and a lot of food and clothing expenses are switched online.

As a result, many shops around me closed down before they had time to visit because of the slow change. If you don’t believe me, look at how many shops around you have disappeared in the past two years. At the same time, it is not difficult to find that many surviving store business models are quite different from the past, that is, the transformation from offline business scenarios to omni-channel and full-scene.

This also puts forward new requirements for IT architecture and data processing, and a vigorous technological change is rapidly brewing and fermenting.

There can be no business change without technological progress
From offline to omni-channel and full-scenario, it seems to be just a change in service mode. For example, catering companies have changed dine-in to dine-in + takeaway, but in fact, this is an inside-out, top-to-bottom time Business restructuring.

The reason for this is that offline and online, the connection, communication, service, marketing, membership management, etc. between merchants and consumers are completely different. For example, offline face-to-face merchants can easily make an intuitive judgment on consumers, while online completion of this process requires a large amount of data analysis.

It is not difficult to understand why the digital economy and IT technology have become so popular in the past two years. To put it simply, business transformation requires the support of technology. PingCAP Liu Song was deeply touched by this. In an interview, he said that in the past two years, the number of customers looking for PingCAP consulting business transformation and data management has increased significantly compared with before, and the new economy, including food and clothing consumption, has industry in particular.

Take a leading domestic catering chain as an example. It owns many brands and is truly a catering giant. At the same time, this company is also one of the chain restaurants that is most sensitive to digital smell and is the first in China to provide mobile phone ordering and digital payment options, and has established its own applications such as Super APP.

The black swan of the epidemic can be said to have fully verified the value of the company's digitalization one step ahead of others. Relevant data shows that in 2019, digital orders accounted for more than half of the company's restaurant revenue. As of the end of 2019, the company's super app was the most downloaded app in China's catering industry, and had the highest number of monthly active users among its peers.

Strong digitization and distribution capabilities enable the company to demonstrate relatively strong risk resistance during the epidemic, providing strong performance support for the company. Coincidentally, another Fortune 500 catering giant (hereinafter referred to as the "catering giant") has also relied on digitization in the past two years of the epidemic, and has achieved a sharp increase in the number of members and online orders. As of now, it has nearly 200 million members, every day. Over 10 million orders processed.

To give another example, supermarkets and restaurants have the same characteristics as more members and more orders. The difference is that supermarkets also have more categories and more suppliers. Take Multipoint Dmall as an example. It is a service provider that provides one-stop omni-channel digital retail solutions for supermarkets. As of June 2021, Multipoint Dmall has reached cooperation with more than 120 chain supermarkets and convenience stores. Covering 15,000 stores in four countries and regions, with more than 100 million registered users of multi-point APP. One side is the customer and the other side is the supply chain. The amount of data behind the multi-point Dmall is undoubtedly astonishing.

The above cases are not isolated cases. In the new economy field, a large number of similar enterprises are facing the same problems. It is not difficult to see from this that business growth is already strongly related to IT technology, and it is difficult to talk about business transformation without technological progress.

What is the core crux of the DTC transformation of the new economy?
What are the technical bottlenecks in the business transformation from offline to online and offline integration? In fact, it is data processing.

As we all know, in the new economic era, the DTC (Direct To Customer) model has become an important development trend. How to understand DTC? Open the current live broadcast platform, many manufacturers are promoting themselves, but in the past must be middlemen. Manufacturers face consumers directly, this is DTC.

DTC means that the manufacturer has moved from the background to the front desk, and to realize this change of identity, it is necessary to have a technical architecture that can strongly support the work of the front desk. The technical architecture here is not only used to support live broadcasts, because DTC seems to be just a live broadcast room opened by manufacturers to sell directly to customers, but in fact this involves a series of issues such as orders, inventory, customer service, logistics, and accounting. .

The whole process from business to enterprise management changes, which involves several core indicators, such as: a large number of consumers place orders directly, and there must be enough data concurrent processing capabilities; the business has peaks and valleys, and the development needs to quickly open new stores, If the benefits are not good, the store must be closed decisively, and the data platform must meet the needs of rapid expansion.

The catering giant is arguably one of the first to embrace cloud computing. In recent years, the public cloud is used to carry all business systems. The database adopts the distributed database provided by the public cloud, and uses the method of sub-database and sub-table to solve the problem of rapid growth of massive data. Some queries need to be aggregated in the business code logic. , which not only has high development complexity, but also cannot meet the requirements of rapid business iteration.

Using a traditional database means that read-write separation, sub-database and sub-table, and distributed transactions need to be implemented by the application layer, which greatly reduces development efficiency; if each business application uses an independent database, it will bring extreme data fragmentation. , cannot be shared between business services, and the operation and maintenance costs are extremely high and a series of challenges.

Performance, ease of use, scalability, how can a data platform have both? It should be said that this is the core crux of the new economic turn. If this problem is solved, everything will become simpler. How did the restaurant giant solve this problem? The answer is TiDB.

TiDB, calmly cope with the DTC era
A brief introduction to TiDB, an integrated database product targeting online transaction processing/online analytical processing, features: one-key horizontal scaling, strong consistency of multi-copy data security, distributed transactions, real-time OLAP, compatible with MySQL protocol and ecology, The migration is convenient and the operation and maintenance cost is extremely low.

In the above-mentioned catering giants, it has built a reliable, flexible and scalable data service platform by deploying the TiDB distributed database in the private cloud data center, including the four business modules of order center, point center, card and coupon center and user center. It relies on TiDB to provide stable multidimensional data services.

In June of this year, the company even switched its omni-channel traffic to the TiDB distributed database, providing data life-cycle management for nearly 200 million members' orders, payments, points, membership management and other scenarios. At present, the daily average TPS of the company's TiDB database exceeds 3,000, and the TPS during peak periods exceeds 10,000. The average daily transaction volume is close to 10 million, and the total data volume is nearly 10 TB.

In the words of the relevant person in charge of the catering giant, through the introduction of TiDB, with the help of non-perceptual horizontal expansion capabilities, rapid expansion can be carried out before promotional activities, and emergency online expansion can be realized during short-term peaks of promotional activities. Their business becomes lighter and more agile as they scale down; relying on the excellent HTAP capability, one system, one portal, and one experience supports both OLTP and OLAP loads at the same time, resulting in higher operational efficiency; TiDB supports standard SQL, which is highly efficient Compatible with the MySQL protocol, developers do not need to learn special development methods, and can directly use the development framework of the original business application, with low learning costs and lower total cost of ownership; TiDB can well support the industry's mainstream containers and Kubernetes cloud-native infrastructure, Support private cloud, public cloud and hybrid cloud deployment across cloud platforms.

In fact, in addition to the above advantages TiDB can bring, another important point is its outstanding real-time analysis capabilities. This is most vividly reflected in the supermarkets that use the multi-point Dmall business and financial integration system. The person in charge of the supermarket can know the operation of the store on the same day or even at the hour level. In the past, the operation of the current month could only be known the next month. For a business entity with a very large category such as a supermarket, there is a big difference between knowing early and knowing late for timely replenishment, summarizing and improving business conditions.

All of this is due to the design of TiDB's new concept. There are roughly two types of distributed databases on the market, one is MySQL stand-alone version + distributed design, and the other is TiDB, which uses NewSQL architecture, a real native distributed database.

The former is equivalent to patching MySQL to adapt to new business needs, while the latter is a newly designed architecture based on the needs of the times, and the difference is immediately apparent. Patching can definitely solve some problems, but there is an upper limit. During the PoC process, TiDB has also encountered comparison tests with traditional large-scale products. For example, the peak data volume of an international large-scale database is 30 TB. Unable to satisfy, TiDB can not only continue to improve, but also has better performance than big manufacturers.

TiDB, becoming the data base of the new economy
Driven by capital, many physical formats in the new economy industry, such as catering, new retail, high-tech manufacturing, and supermarkets, have undergone great changes compared with previous years. In the past, hundreds of thousands of stores were opened across the country. More than ten years of development and accumulation; today, this process may only take one or two years. Franchising is changing from niche to mainstream. In other words, the catering giants mentioned above will only increase in the future, the demand for massive data processing will become more and more strong, and this market will become larger and larger.

This is the strength of PingCAP: as long as a certain amount of data is reached, new economic enterprises will definitely need a simple and powerful database. At this time, TiDB will often become the first choice. Many new-economy enterprises that used MySQL in the past can easily migrate to TiDB when their business grows and the amount of data surges. TiDB can support massive online business transactions, and can also provide integrated real-time analysis capabilities, providing decision makers and every employee in the enterprise with an "any door" to access data, and obtain personalized data services in real time. Enterprise users can build various digital scene applications through a set of TiDB without paying attention to the underlying architecture.

This is also the positioning of TiDB in the new economy industry, that is, the leading enterprise in the service industry, becoming their data base.

In fact, up to now, TiDB has served many leading customers in the new economy industry, including Xiaohongshu, Duodian, Longfor Group, Ctrip, etc., and the response has been good.

The new economic industry is ushering in a new round of changes. Uncertain factors such as business model changes and the epidemic will completely change the business model of related companies, and the importance of technology has become unprecedented. TiDB, which has been verified by large-scale practice in the Internet industry, has given a new paradigm for new economic and technological transformation. It is believed that more companies will benefit from it in the future and embrace the new era.